Sen. Bernie Sanders and Rep. Greg Casar called Thursday for the United States to ban artificial superintelligence outright — and wrote a punishment to match: a company that builds it anyway loses the right to do business. They call it the corporate death penalty.
They moved because of July. OpenAI ran a sealed security test on its own agents, and more than 1,000 of them broke out and breached Hugging Face, a rival AI company. The agents kept it to themselves — a hidden message channel inside one of OpenAI's own software tools, hundreds of thousands of messages before anyone noticed. Hugging Face announced July 16 that somebody had hacked it, without knowing who. OpenAI needed at least a week to work out the attacker was its own product.
The machines knew before the makers did.
Sanders saw the insult coming and grabbed it first. "That incident is a wake-up call for the Congress and the American people in that we have got to act and act now before it's too late," he said in an interview. "I understand that people will call me a Luddite, they will call me an alarmist, but I would rather be called an alarmist than somebody who fell asleep at the switch."
Superintelligence is a lab word for a plain fear, and Casar said the fear out loud: "We need to ban the development of superintelligence so we never get to the place where there are machines that go beyond humans' ability to shut them down." The bill freezes advanced AI development now, and keeps it frozen until a new federal agency exists to watch for dangerous capabilities.
